@Play 80: Welcome back to the Dungeons of Doom

You are lucky! Full moon tonight.

This marks something of a revival of my old GameSetWatch column, @Play. That one slowed to a halt because it was getting harder to avoid repeating myself, or had to resort almost to filler columns to keep it going. Particularly, I had a sense that I had said everything that needed saying. That's not a bad excuse to not say anything.

But then something happened. Like a gas spore struck by an arrow, roguelikes exploded.

Before, the closest to mainstream roguelikes were things like the Mystery Dungeon series, which were a stretch even in their native Japan, or the Diablo games. Now, it seems almost like every other new indie game on the Steam store is tagged roguelike, 108 of them as of this writing. Before our hiatus, Spelunky (one of the best real-time roguelike-inspired games) was a promising freeware creation. Now it's available for Xbox 360, PS3, Vita, and Steam, has fascinated hundreds of thousands of players with its terrific procedurally-generated gameplay, and has been the focus of many livestreams and YouTube recordings by star players like BaerTaffy and Bananasaurus. That indicates, to me, that the lessons of roguelike games have gotten out to some extent, and even been embraced, and I find that heartening. And ToME, under the name Tles of Maj'Eyal, is there, and ADOM is coming, and Desktop Dungeons has been there for a while!  But it's not enough.

Anyway, there is no reason to restate the things I already said well enough on GameSetWatch. That site is still up (thanks Simon!) and all my columns, for their faults, are still there, and most of the things I said then are just as valid now. But it might be good to recap the basic points I made during its run. Here are most of them, in summary, and after that we can move on to new material.

All About @, Again

Possibly of use: The first column of the original run, from eight years ago, which contains a different, shorter, introduction.

rogue (from package bsd-nonfree)

A roguelike game is, in the broadest sense, a game that is like Rogue. One way to stylize that name, by the way, is with a lowercase first letter, as "rogue," as it is a Unix command. But you can be forgiven for writing it like Rogue, and that's what I'm going to do. (The same goes, by the way, for nethack/Nethack, even though that game is officially called NetHack, and I'm going to try to do better about it.)

There is no official definition for "roguelike," nor an official body of any kind who could define it. It was adapted by fans to refer to games that took Rogue (1980) as a starting point and elaborated upon it. It itself took inspiration from the Dungeons & Dragons books (OD&D: 1974, AD&D 1E: 1977) and possibly earlier computer RPGs like dnd and Oubliette, which can still be played on the PLATO system running on Cyber1. Some of the earliest roguelikes, after Rogue, are nearly forgotten games like Advanced Rogue, SuperRogue, Ultra-Rogue, and Xrogue. It is very hard to play them in their original forms, but modern ports of some can be gotten from the Roguelike Restoration Project. Other prominent early roguelikes include Moria, Larn and Hack. Some of those games got their own variants: Moria was modified into UMoria and then Angband, Larn to ULarn, and Hack to NetHack. New variants are produced to this day. Angband has dozens of variants, and NetHack has gained quite a few just in the past few years. There's also standalone roguelikes ADOM and Brogue, among others. All of these are console-based, text games, but it is not necessary to be a console game to be roguelike, especially not now.

Rogue is one of the most interesting games, relative to what was floating around at the time of its creation, that there ever has been, and it's still pretty fascinating now. It was released the same year as Pac-Man. I refer to it in the present tense because people still play Rogue. It is a randomly-generated exploration game, using a simplified version of the original Dungeons & Dragons rules, in which the player guides a surrogate character through a treacherous labyrinth of monsters and traps. Rogue is particularly known for its great difficulty, its system of permanent character death (usually shortened to "permadeath"), its variety of play experience, its random game world, its tactical combat, the deep nature of its simulated world, its unique item identification system , and its replayability. Many of these things individually have given a game reason to be called, at one time or other, by some person, "roguelike." Because there is no official definition of the term, despite some creditable efforts like the Berlin Interpretation, lots of people use it to describe lots of different things, so let's briefly address each of the things in that list.

Difficulty: Rogue is very hard, and some later games are even harder. Some versions of Rogue are particularly hard.

moria (Linux Mint)

Rogue belonged to what I continue to call the "Mount Everest" school of design, where the game is supposed to be a just barely surmountable obstacle, to a hypothetical player. You may not win this game now; you may not win it ever. But even if you lose, if the game is doing its job right, it feels good if you just play better than you usually do (this is why many roguelikes provide high score tables). Or at least, it should ideally. The game is very hard, but you know it's very hard, so failure is not a matter of shame. If you're the kind of person who absolutely must beat every game you pay then roguelikes may not be for you.

But difficulty isn't enough by itself. There are lots of hard games that aren't really roguelike in nature, indeed most games released to arcades. Difficulty is by no means exclusive to roguelikes, but in subtle ways improves many of other other aspects here: if the game isn't hard, then why do you care about exploring unknown dungeons or identifying items?

Winning at Rogue is something to aspire to, with practice, experience, and (sometimes) lots of spoilers. But even with spoilers, winning is far from guaranteed, and for this fact looking up FAQs is considered to be fairer game than other, more static genres. This is because of--

Randomness: Every time you play Rogue, the game generates the dungeon anew.

But there are lots of games that use randomness but aren't roguelike. Particularly there are games that use it in a trivial way, that is, in a way that doesn't really affect the play. If you randomly generate a dungeon layout it may seem rogue-ish, but there are lots of ways you can lay out rooms and corridors that don't ultimately make any difference to the choices you make, not if you always stock that dungeon level with the same items and monsters, not if the critical path is always functionally the same, not if there isn't some resource management aspect to exploration that gives weight to that randomness. And it's also possible for a game to be too random, where the spikes up and down on the player power graph matter more than the player's skill. Usually good roguelike design is "spikier" than you might expect from playing other games, but still, if you're constantly running into top-of-the-line Dragons on level one, it's possible that the game is more about hoping the random dungeon generator doesn't generate impossible situations. That's especially important because of--

NetHack 3.4.3 (console, Linux Mint)

Permadeath: When you die, that doesn't mean you load in your last save. The game is over. The next game, you must start from scratch.

This is still the most controversial aspect of roguelikes, although it really shouldn't be. Games were like this from the very early days. You can't save your place in Space Invaders and return to it. Likewise, if your character dies in a Dungeons & Dragons adventure, at least at low levels, he's gone, your DM isn't going to resurrect him on a whim. At least, not if he's any good.

The common defense of permadeath given by enthusiasts, which I've been known to use myself sometimes, is that it gives weight to your decisions. But I think that's kind of a cop out. This isn't a special attribute of roguelikes, that's just what games are, it speaks more to a deficiency in modern gaming that gives players unlimited do-overs. It was a niche idea to be able to return indefinitely to saved positions, originally used for adventure games where exploring the consequences of decisions is part of the fun, that snuck out of its genre into other genres where, divorced of its context, it lost its original purpose and became something games just had to have, like happened later with experience levels, character unlocks and loot grind.

One variation on this idea I've heard bandied about lately has been called perma-consequence, the idea of there being other states than death that might not be simply reversible. Rest assured, we'll be returning to this idea.

Variety of play: This is not a commonly recognized attribute, but it gives a game a kind of roguelike feel. Random dungeon maps by themselves are not enough to make a game substantively different between plays. The player must be offered meaningfully different choices between games. In Rogue, depending on what items you find, or which enemies end up being most common, you may end up playing in an entirely different manner. For success, the player must to some extent adapt to the situation, rather than adapt the situation to his playing style.

This is such an essential part of the best roguelikes, yet it's recognized so little, that I'll probably be devoting a column to this one before long.

Tactical combat: The classic roguelike paradigm involves exploring a grid-based world where enemies have largely the same abilities as the player. Movement and attacks each take up an amount of time, and so one must frequently make decisions as to the best way to fight, or whether to retreat.

I'm not devoting that many words to this one because the meaning of roguelike has been drifting as of late. I'm still something of a die-hard, the word to me still tends to mean Turn-based Overhead Tactical Combat (Am I allowed to coin spurious acronyms, like "4X"? If so, how about TOTC, pronounced tot-ic?). The term roguelikelike has been suggested, but it puts me to mind of a certain shield-eating Zelda enemy. I suppose you could also quasi-roguelike (because "quasi" is a cool prefix) or rogue-ish. Just so you know, we'll be covering partly roguelike games too.

Depth of world: Rogue's game universe supports many different kinds of action, some of them useful only in obscure circumstances, but everything is useful at some point or other. Walls and floors can be searched for traps; you can rest in one place to regain hit points faster; rings provide special powers but at a cost of food consumption; you can throw arrows sure, but if you take a turn to wield a bow first they'll be much more effective; armor greatly increases your survival odds against powerful monsters, but wearing it makes you uniquely vulnerable to the one monster who cannot inflict physical damage on you; and the mere act of dropping a certain kind of item can save your life. Many actions turn up unexpectedly useful, but only experience (or spoilers) can tell you how.

If you consider depth of world to be measured in the variety of things the player can do, then this could also be considered one of the hardest hurdles to clear in learning how to play Rogue. The keyboard interface could be considered obtuse to present-day players. Its key layout is heavily inspired by the Unix text editor vi, using hjkl for orthogonal movement and allowing players to prefix some commands with numbers in order to repeat them (like 20s to search twenty times in place).

Angband 3.3.2 (ASCII window, Linux Mint)

At the time and places of Rogue's initial popularity most players could be presumed to know vi keys, and be used to its command mode which is the inspiration for Rogue's terseness of communication, but few people these days who aren't Unix mavens learn vi. (If you're interested in learning vi, playing Rogue could be considered a way to get used to its movement system, although its diagonal keys are not used there.) By the way, most modern variants of Rogue and console roguelikes support using a number pad for movement. If you're interested in playing classic roguelike games on a laptop, you should either make sure to get a model with a numpad or invest in a cheap USB numpad keyboard add-on. (I got mine ten years ago and it's still kicking.)

Anyway, even if you know vi, there's plenty of new key commands to learn. In most cases these commands are pretty obvious by their letter, or are not too dangerous if you press keys to figure out. To eat something you press 'e', for instance, and to read something press 'r'. A few are more idiosyncratic: to drink a potion, you press 'q', for "quaff." 'w' stands for wield and is how you equip weapons; Shift-'W', however, means wear, and is how you put on armor. One thing that's nearly universal to all roguelikes, however, is 'i', for inventory.

Item ID: This is one of Rogue's less-copied features, but one that has a profound effect on its gameplay. As other games hosting random dungeons have increased in number, this has become probably Rogue's most defining characteristic, and the one enamored-of most by the Hack-like branch of its descendents.

Scattered throughout the dungeon are randomized magic items going by different descriptions, like "orange potion" or "lapis lazuli ring." Their purposes are scrambled within their item classes from a number of possible effects that are consistent for their descriptions. An orange potion, for instance, may be extra healing one game, gain level the next, and blindness in still another. Importantly however, while their functions change from game to game, within a single game they're the same: orange potions won't suddenly take on a different quality but remain consistent in function until you die. (This is one reason that permadeath is important to Rogue.)

Mixed in with the good magic items are some bad ones – drinking a potion of blindness is generally a pretty bad thing to do in Rogue, but because you don't know which kind of potion will blind you before drinking, you usually get burned by each type once, at some point. This can be escaped, however, through the use of a special kind of item, the scroll of identify, which will infallibly identify one item you're carrying. Unfortunately, scrolls are themselves one of the random item types, and so you will have to use-test at least a few items before you find out what they are. (I call use testing items to find out what they are "trial ID," as in trial-and-error. I'm not sure why I started calling it that; maybe it reminds me of drug trials.)

Once the game is satisfied that you definitely know what an item is, it will rename it for convenience: it'll stop calling the orange potions and instead say something like "potion of extra healing." Many roguelikes that adopt an item identification system, especially the more difficult variants of the Japanese Mystery Dungeon games, will identify these items immediately on their first use, under the philosophy that if you took the risk in using it you deserve to know what it is, but Rogue does not always do that. Instead, if you use an item, it will be identified for you if the game judges its effect was obvious enough. Sometimes this is nearly always (it's usually pretty obvious when you've been blinded), but other times may be situational. A few kinds of items will never be identified from use, but clues may be provided upon use. Sometimes the game will ask "What do you want to call it?" after using something, asking you for a name. It will then rename the item to that, saying, depending on what you entered, "a potion called extra healing." You can also manually name items with a special command.

By the way, Rogue's item identification system most likely has its root in 1st edition Dungeons & Dragons, which has a number of cursed and otherwise-bad magic items that exist mostly to be confused with good items of the same type. In AD&D however, many of these items were outright lethal, like the Bowl of Watery Death, which existed only to screw over players who thought they were getting a Bowl of Commanding Water Elementals. Rogue versions, at least, are much fairer: no Rogue magic item can kill you by itself, although some may end up deadly combined with other situations.

Dungeon Crawl Stone Soup 13.1 (console, Linux Mint)

Replayability: This comes last because it's the most important. All of the previously mentioned features are in Rogue at the service of this one. Rogue is a game that's intended to be played many times. Most of the time you'll lose, but even when you win, it'll still be fun to play next time. This is the primary measure, in my eyes, of whether a game that calls itself "roguelike" is doing justice to the fearsome legacy it invokes. If you can't start a dozen games and still feel interested in playing from the first move, then there's an important lesson there that hasn't been picked up.

A game where you can't die, but is varied enough that each game feels like its own adventure, is, in one sense, more roguelike than a boring game with permadeath, random dungeons and scrambled magic items.

Here are some other relevant points I've argued for over the years:

• While trends and fads may make some things seem more interesting to us, at different times, than others, game design doesn't go obsolete.
• Grind is bad; players should not be kept from the "good stuff" in a game by having to pay their dues in terms of time played.  This is degenerate game design.  Think carefully about the consequences of this; while there are fewer always-bad things a designer can do than you might think, as a designer, you know why you're doing it.
• To avoid becoming grind, player improvement must come with risk of loss, either of resources or outright game failure.
• While permadeath can be good, preserving state in a game long-term, between games, is a very interesting idea – it's what gives a game continuity of experience, that makes a play one long game instead of string of isolated, unrelated situations. There is probably an upcoming column on this matter.
• Randomness can be very interesting in a game, if used properly.  However, random dungeons by themselves are not enough, if the contents of those dungeons is always the same.  You might as well just have pre-made dungeons in that case.
• Pure randomness can be worse than pre-written content; at least the latter is presumably playable once.
• Good random content generation filters its randomness, uses it to create good content while remaining both challenging and winnable.
• That said, good roguelike design doesn't take it easy on the player based on his state.  If you're low on health, sources of healing do not magically become more common.  This is because roguelikes are fundamentally exploration simulations, and cooking the dice in favor of the player is breaking the simulation.  Roguelikes are not play experiences, they are challenges; you don't play them to feel empowered, but to see what will happen.
• Longer games must be more fair to the player, must be more certainly winnable, and must generally be easier. For all these reasons, shorter roguelikes tend to be best. The sting of a loss is proportional to the amount of time spent playing.
• Balancing this out, winning a good roguelike game does not make future wins foregone conclusions. A well-designed game can still be challenging and fun even after many victories.
• The first time you finish a good roguelike, you generally have the feeling that you lucked out. You probably did, but you learned a lot to get there.
• Tactical combat is considered by some to be an essential element of roguelikes, but increasingly many interesting examples of the genre do not require it.
• Some roguelikes have interesting features by which the events of successive games can modify the game world, and affect later games. At best this results in features like NetHack's bones levels and Shiren the Wanderer's multi-game quests; at worst, the player may end up having to "level up"* the game world to make it easier in order to have even a chance at winning. We'll talk about this more when we get to reviewing Rogue Legacy.

Where are we going from here? Well, there's a lot of game design issues regarding roguelikes left to cover. Then there's covering tournaments and game jams, including the always-amazing 7DRL which is going as I speak, and may become the focus of the next column! There's reviews, and lots of them. Variants and patches to tell you about. Interviews and follow-ups. The occasional playthrough. Trivia, novelties and oddities. Sometimes the work of other enthusiasts. That is, all the stuff we used to cover. I am, however, expanding the focus of the column somewhat. Roguelikes are still the primary focus, but because the roguelike flavor has leached a bit into other genres, we'll talk sometimes about those games. Roguelikes are one of the prime examples of what bigshot game devs call "procedural generation," and I'll take that as my excuse to talk more about those issues, and the philosophy behind it.

This is nothing that long-time readers should look askance at. If you are such a reader, welcome back! If you're a newcomer, don't worry! All new readers receive a pile of food and some darts completely for free, and one complementary expensive camera. Be sure to take lots of pictures of the wildlife; your life may depend upon it. When (if?) you see Twoflower tell him I said "Hi."  Until next time, once again, this is John "@rodneylives" Harris, still waiting by those down stairs, wondering what happened to that dog.

* A final note. This is going to sound pedantic but I don't care, I spent years writing @Play and figure I've earned some orneriness. I insist that the proper term to use in roleplaying games, if you have to use experience levels at all, is to "gain a level," not to "level up," which is Engrish come home by way of console JRPGs. I am playful with this insistence because I don't actually care.  But if I were ever to edit a New Yorker-style magazine somehow devoted to video and computer games, then this would be the first thing I threw into my gnomish and inscrutable style guide. And diaereses over adjacent vowels in separate syllables, of course: this is a civilized column.

All screenshots taken in cool-retro-term, running on current Linux Mint.  It's the best way to pretend you're back in the 80s playing on a flickering CRT!

--------------------

@Notes:
As said above, the 7DRL Challenge is going on right now, where a bunch of guys try to create a roguelike game in a week.  A few important roguelikes, like DoomRL, got their start as 7DRL jamgames, and it's notable because every year there are both an unusually high number of both entries and awesome projects.  You should check it out!  One Twitter user who's been following them is @dungeonbard.  Maybe worth checking out?

@Play 79: The Re-return of ToeJam & Earl

This is the beginning of the revival of @Play, my old roguelike column, formerly of GameSetWatch. I had intended for the first column to be a retrospective over the length of the original run. But seeing as how there's over seventy columns in the sequence, that's taking a bit longer than I expected, and something came up. Something pressing. Something that demanded that I write about it, here, first. Sorry about that, next time will be the "official" relaunch, whatever that means.

What could be so important? People who remember the first run of @Play will remember that one of the earlier columns was entitled ToeJam & Earl, the Roguelike that's not an RPG. I made my case, I think pretty well, about the inclusion in the genre of a game that's not a tactical, turn-based D&D-styled fantasy combat game. and I think the case is even stronger now that games that fall even further afield are getting that label stuck to them.

ToeJam & Earl is a game that is dear to my heart. Although it is a slow-playing Sega Genesis game with a surprisingly low framerate in normal situations, and many other weirdnesses besides, it has excellent design and gameplay. Its two player co-operative mode is possibly the best of the type ever realized. Its characters, even the enemies, are rendered with a cartoony charm that doesn't get old over many replays. But that's okay, because TJ&E is designed to be played again and again, and even after being won it's still an interesting game to play. Because, as you can probably tell from the title of the column I mentioned, indeed the fact that I'm bringing it up here, TJ&E has many roguelike characteristics: randomized items, random maps, and yes, the ability to lose the game too. And because TJ&E is a very challenging game, players will most likely lose their first several plays, before they learn the many tricks necessary to survive.

When I first played it I had played Rogue before so I could see the design relationship between them, but I didn't know of many of its successors. Hack and Moria were not on my radar. I might have read about Larn once (in a review of an Amiga version in the pages of Commodore Magazine). NetHack was around, but I think it was another year before I found out about it. In TowJam & Earl I found a game that had many of the addictive qualities of Rogue, which I knew of directly from playing it on a relative's DOS PC, yet it was on my Sega Genesis! And what's more, if I had a friend over, we could play it with each other! It probably remains my favorite Genesis game of all, and that's despite some strong (non-roguelike) competition.

ToeJam & Earl wasn't the first console game to borrow enough elements from Rogue to make it eligible, in the main, to be called a roguelike. While Chunsoft's Japanese series Mystery Dungeon didn't make its Super Famicom debut until 1993, Sega had developed a roguelike called Fatal Labryinth for its short-lived SegaNet service, a few months later produced on cartridge the similar Dragon Crystal for its Game Gear portable system, and eventually made a cart version of Fatal Labryinth. Those games are merely okay, though.

I liked what I did for the original TJ&E column above, and it continues to stand I think. To get to the point, I bring it up because half of the team who designed the original game, Greg Johnson, said in an interview with Venture Beat that he plans on doing a Kickstarter to fund the creation of a true, in-the-spirit sequel to ToeJam & Earl. This is big news, although it should also inspire some wariness. To explain why, I will have to delve a bit into the history of the original game's reception, and why it hasn't really gotten a full game in the mode of the original in 24 years, especially since that's what the third game was supposed to be....

I. The Wake of TJ&E Genesis

Here's a synposis of the original game. The heroes, the titular ToeJam and Earl, are two aliens from the planet Funkotron, which is kind of like a more cartoony and kid-friendly version of a P-Funk show. Funkotronians are a happy and peace-loving lot, devoted to music. While cruising through space in their (peers down nose through glasses at script) "Righteous Rapmaster Rocketship," Earl steers their capsule straight into an asteroid. The machine crashes into ten pieces scattered around the least funky planet in the universe: Earth. To return home, the two aliens (or just one of them if you're playing solo) must explore the planet and reassemble the ten pieces. However, the Funkotronian perspective on spatial physics must be radically different from those of a native of the Earth, because what appears as a round planet to us is presented in the game as a set of 25 floating platforms hanging vertically in outer-space above each other. Each of these "levels" bears somewhere on its surface a flashing yellow elevator. Entering that takes the player up to the next level of the 25. One piece is always hiding at the top.

You go up by intentionally taking the elevator, but often you end up going down involuntarily. If you step, jump, plunge, or are thrown or knocked off one of the levels, you'll end up at a random location on the level below, and have to re-find the elevator to get back up. Compounding the difficulty of your search, the levels are randomly generated each game (but persist during the same game) and are full of Earthlings, of varying degrees of malice. Most Earthlings have some kind of attack they can perform that damages you, but many have, either instead of or alongside a damaging attack, some special property that can cause you problems. Hula Girls randomly inspire your character to hang out and dance.  Cupids can mess up your controls if they hit you with an arrow, Moles steal items, and Boogeymen are usually invisible. Both the number and challenge of the monsters increases as you ascend through the levels. In true roguelike fashion, survival for any length of time is extremely difficult at the end, so you have to move fast, get what you need, and get out, and probably consume some of your items to do it safely.

Unlike most roguelikes however is one of the most interesting aspects of TJ&E's gameplay: the players are defenseless by default. In normal situations, they have no attacks. You never have to kill enemies during the game, but there are situations where they might be directly blocking the way across an essential path, or are just pursuing you closely and refuse to give up. Fortunately, the duo can use presents, which I suppose are remnants of a huge messy birthday party, they find scattered around the levels. There are 25 different kinds of these, and they're analogous to the potions and scrolls of Rogue: once you use one it takes effect immediately.  There are both good and bad types, but you don't know what they are at the start of the game. They're saved for you in a limited inventory screen.  Once you know what one of a type looks like, all other presents with the same wrapping paper will contain the same thing, and the game will even remind you of this information. You can pay a random wise man (naturally found dressed in a carrot suit) to identify presents for you using "bucks," which are also found scattered around.

There are a wide variety of these presents, and discovering them is a major part of the game. Some contain Tomatoes which can be used as missiles to turn the tables on your pursuers, or a tomato Slingshot, which is even better. You can use Super Hi-Tops help one zoom away from them, Spring Shoes to soar over their heads, and Rocket Skates to rush away in a roar of jets that's just as likely to plunge you down a level or two—oops. There are presents that heal, reveal some of the map, and award an extra life. There are also presents that put you to sleep, summon a damaging raincloud, or even instantly kill you, but fortunately Funkotronians get multiple, although not infinite, lives to live. The very worst present of them all, even worse than the instant death present, is the diabolical Randomizer, which randomly scrambles all the presents in the whole game, forcing you to have to start learning what everything is anew. After you've identified 15 of the 25 without finding one, opening unknown presents gets progressively more and more tense; the strength of the game's opposition is high enough that an unfortunate Randomizer opened on level 20 can doom your game, even if you're in great shape otherwise.

This isn't a full description of the game's many charms. Here is a YouTube playthrough of a successful single-player game of ToeJam & Earl, played in Random Mode (the way you're supposed to play it). It starts a bit slow, but the first present he opens is the instant kill Total Bummer, and at about 21:30 he is stumbled upon by The Randomizer.... Here, the Game Grumps fight through a few minutes of the game in two-player mode, in their cheerfully dismissive way.

II. Public Reaction & Panic on Funkotron

Most console games, even back then, sell strongly at first then trail off over time. ToeJam & Earl reversed that: it started slow, but built up as time passed and people found out how witty and fun it was. Let me assure you, who has seen witness to many worthy games without ad budgets to speak of fade silently away, unloved and unmourned, from the shelves of stores. What happened to TJ&E wasn't supposed to happen. They broke from the script.

And even if you played it for ten minutes, so that you can assure your friends that you "gave it a chance?" You can tell watching the Grumps on their play through, they're not seeing it in their early run. They don't even open a single present. The game has to grow on you, and most people aren't looking for that now, and weren't looking for it then either. But it did grow on people, I don't know how but some people besides crazies like me gave it a shot, through what sufficed for spreading "virally" before the Internet, that is to say, through plain old word of mouth, one addicted player showing it off to his friends, and going on from there. Even now the cult of the game lives on. (Aside: how did Sega choose to market this game to people? Here's a TV ad for the original game. Here's one for the Genesis sequel. Er, yeah.)

I can't tell you how it caught on. I can only tell you what appealed to me about the game: the challenge, the humor (part of that being the digitized speech), the charming characters, the mysterious elements, and the replayability. I assume that people saw in it what I saw in it. And I say, anyone else who tells you why the original succeed or failed is also making an assumption, unless they've conducted actual polls of its players from around that time, and even then I will have my doubts.

Whatever it was really, Sega decided its design elements, its roguelikeness, couldn't be it. That's why Sega asked creators Greg Johnson and Mark Voorsanger to abandon their efforts at making a sequel in line with the original, and make a more traditional platformer for a second game. (Note: speaking honestly, this is really the third TJ&E game. The true second, called Ready, Aim, Tomatoes!, was a mini-game included in the pack-in cartridge for Sega's Menacer light gun accessory. I have not played it, but everything I've heard about it suggests that it is slight. If you want to be pedantic, I think the Art Alive cartridge has a few TJ&E character stamps on it too.)

A complete description of the sequel, ToeJam & Earl: Panic on Funkotron, doesn't fit in with this column, because it is not a roguelike at all. There's no inventory, no tactical positioning, and no randomness. It is fairly challenging, but it's the same challenge every time. Some of the humor is back, and the game play is quite polished, and truth be told it's not really a bad game at all, though surprisingly long. Two-player co-op is very hard to do in a platformer, and it does a good job. It has large, well-animated and appealing characters, vivid, colorful graphics, and huge worlds to explore. But fans agree, it's not the same.

III. ToeJam & Earl III: Mission to Earth

If the story had ended there, well, who would have blamed them? There are lots of popular games that never spawned a sequel, or just one sequel and no more than that. Something about TJ&E inspired another game nine years later, for the original Xbox, called Mission to Earth. It didn't do well in sales, despite a front-page story on Play Magazine. (Do you remember Play Magazine? I do. Sorta.) I don't have sales figures to hand, but the reviews mentioned on the game's Wikipedia page range from moderately good to quite poor. The bad ones sometimes cite good reasons (the dated hip-hop theme displacing some of the original's whimsy), and sometimes bad ones (one mag claims its random maps necessitated generic missions, when a game like this shouldn't have had missions at all). In short, like many also-ran games that have a glimmer of greatness hidden within them, nothing really to distinguish it.

TJ&E III was supposed to mark a return to the style of the original game, and was touted as such in marketing, probably because someone at Sega remembered the failure of the original sequel. It returned in some ways, but didn't in others, and as I've said before, a game can be roguelike in many ways except for one thing, and that one thing will wreck the whole design. Take permadeath out of Rogue, or its challenge, randomness, variety of monsters, variety of items, clever tricks, or food system, and you have a greatly diminished game. All these things support each other; you can't half-ass it. If you take out even one part you had better know what you're doing. It is not a friendly genre to executive meddling.

Reader, I have played this game. Do not laugh, but I owned a copy of this game. I bought an Xbox specifically so I could play ToeJam & Earl III: Mission to Earth. I can tell you what I found disappointing about it. I suggest my reasons are universal ones. I'm sure you could find someone who would offer different ones. I think they're wrong and I'm right, but what else would I say? Anyway, here is what I think the game did wrong:

- The game organizing model is completely wrong. It isn't a short-length game in a with limited lives like the original, but a long game organized as hubs with level doors and with no serious penalty for failure, where each finished level is basically a checkpoint. Its structure is similar to that of Mario 64-style platformers: collect things in order to meet unlock requirements for later levels, from which you collect more things. The game saves after every level, which puts you back on the hub. If you fail, you go back to the hub. It's structured as an unloseable metagame wrapped around smaller level-length games, completely at odds with the tension and danger that soak the original's later levels. In short: no permadeath.

- The default mode is "Story Mode," as opposed to "Random Mode." Story Mode offers a strongly-designed experience with mini-games and missions and such. Random Mode offers a lesser version of that. The Xbox version of the game was obviously designed to be Story Mode first and Random Mode as an afterthought, while the Genesis game was Random Mode for everyone and Fixed Mode for those who couldn't hack it. In other words, diminished randomness.

- The original game's present system returns, with both good and bad presents. But the whole point is lost, because all the presents are known from the beginning. I've written before about how Mystery Dungeon games tend not to be very interesting until the items start to get randomized, which is usually, annoyingly, frustratingly, well into the bonus content. The same is true when the game is ToeJam & Earl. One of the major themes of roguelike gaming is discovering the world as you play, but here all this information is handed to the player on a platter. It is possible for items to be scrambled in a limited fashion later, due to the attacks of Medusa Baby monsters, which act like minor Randomizer items. And once scrambled, the danger of opening Randomizers themselves reappears. But identification is easier in TJ&E III, and with careful play it's possible simply to never be hit by a randomization attack. So to use some jargon, nerfed object ID, and, since bad presents are limited in how badly they can mess you up because of the absence of permadeath, tensionless trial ID.

- The game objectives are more about fulfilling missions than the primary goal of collecting albums. The quest structure makes you spend a lot of time doing arbitrary things as makework to pad out the length; and the length has to be there because the game is basically unloseable, since the play cycle goes between a hub and levels instead of between individual, challenging games. There are games that do this successfully (like Mario 64 for one), but they all have a lot more play variety; each Power Star is a completely different challenge from the one before. Lack of essential challenge leading to unnecessary length.

- The original ToeJam & Earl was a surprisingly pacifist game. The guys have no attacks on their own; all their weaponry comes from occasional random presents, and have time limits. The game is designed so that you don't need to kill anyone, but there's still a lot of enemy opposition to deal with. The Xbox game gives players a melee attack with a good range, and "funkify notes" that can attack at a distance. The latter is limited, but common. To summarize, the player is overpowered relative to the enemies, which also takes further weight off of the item identification system, since the player doesn't need to rely as much on presents to survive.

- This is a digression from our focus on roguelikes but it has to be said: the theme was dated, even relative to its release date.

Even the original wasn't really about what it said it was. ToeJam and Earl's text captions presented a mixture of funk cliches and surfer speak as might be viewed by a fourth-grader, but it worked because it was an extremely silly game filled with Hitchhiker's Guide to the Galaxy style humor. (Seriously, read the Genesis game's manual, which includes illustrations and funny descriptions of all the game's enemies. It's great!) The best thing about Panic on Funkotron is that it kept the original's tone exactly right while managing to fill in the gaps about the duo's home planet. The terrific music was the icing on the cake.

TJ&E III mixed in an intro showing off a truckload of cliches, including some surprising innuendo, for a whimsical cartoon game anyway, about ToeJam's "third leg." It's possible this was added due to (somewhat justified) self-consciousness about appropriating black culture. I'm more inclined to blame Microsoft's desire to differentiate their nascent Xbox console from the Gamecube. Now I admit, I am not a terribly funky person. So, I present this video that contains the game's intro, so that you may judge for yourself. Due to the racial issues involved, I do not know if the best route was to return to the original's fourth-grade concept of funk. The Dreamcast beta that's surfaced doesn't have the intro, and the theming seems less intrusive in it, so it could have been another thing added by Executive Meddling. I have little inside knowledge of this. (I do have a tiny amount, however, from an email exchange I had with co-creator Greg Johnson several years ago. I don't remember much of it at the moment, but may drag it out for a later column.)

Anyway, one way to go could have been to keep the funk in the game's music and the occasional nod, and increase focus on Hitchhiker's Guide depictions of dangerous, unfunky Earthlings. There, that concludes the non-roguelike portion of this essay.

IV. The Funk of the Future

Interestingly, TJ&E III's development started on the Sega Dreamcast, and in 2013 a disk of a playable beta surfaced on the internet, which you can burn to a CD-ROM yourself and play. It can be found without a huge deal of trouble through web searches. Here is video of the game. Here's an Escapist blurb about it. I've not played it yet, but will soon and report back for a later column.

Take particular note: most of the qualms I mentioned above don't apply to the Dreamcast beta. One could take this as indicating great things for the Kickstarter project. But also of note, just remaking the original game probably won't be enough anymore. Of course, roguelikes have progressed somewhat since the Genesis version of TJ&E. There is a world of new concepts to borrow from, in recent games like Spelunky and Brogue, heck, even from Mystery Dungeon and NetHack. Particularly, the SNES/DS Shiren the Wanderer paradigm of a "meta game" that advances slightly as the player plays the main game.

There has never been a better time to be a roguelike player. Or, to be a roguelike developer. The fact that, of this writing, the project Kickstarter is one day old and already approaching half of its $400K funding goal is solid proof of that. Only time will tell if ToeJam & Earl are ready to walk up and stand alongside Spelunky Guy at the forefront of the genre, or if they'll be content to remake (or even wreck for a third time) a 24-year-old game. That said, I think Greg Johnson has a good eye for the game's requirements, and wish him all the best.

Here is the link to Greg Johnson's Kickstarter for ToeJam & Earl: Back in the Groove.

The corporate site for ToeJam & Earl Productions Inc. is still on-line. This is its contact form, although I don't know if anyone still reads it.
Special thanks to Dontae Lawrence for assistance with this column.

----------

@Notes –

Here is my list of upcoming topics. If you have something you'd like to hear about, let me know at johnwh(at)gmail.com, or Twitter handle @rodneylives. There are in no particular order, and presence here doesn't definitely mean I'll get to them. They're just what I'm considering at the moment.

Meta:
@Play Review (next time)
Play & Dev topics:
World generation & backstory
On the nature of exploration gaming
What makes a good monster?
Item design: What's important?
Item design: How much is too much?
How pared down can a roguelike get?
An overview of roguelikes on Steam
Ideas on overcoming the existence of FAQs
The role of skill in turn-based games
Gameplay vs. Simulation
On the nature of "knowledge," "skill," and "wit"
The Secret Competition: why players grind even when they don't have to

Reviews:
Spelunky HD
Rogue Legacy
Auro
ToME & its Steam release, Tales of Maj'Eyal
Darkest Dungeon
Catacomb Kids
Pixel Dungeon

Here are some general game finds:

A free game to look into, I haven't played it yet so I cannot vouch for it, is Pixel Dungeon. It's for Android and Windows, but the Android version has in-app purchases.
* http://pixeldungeon.watabou.ru/.

Another one I've yet to play is Escape From Cnossus, playable via web but also as a Sinclair Spectrum game.
* Game
* Review